Accepted Manuscript Regional ventilation distribution and dead space in anaesthetised horses treated with and without continuous positive airway pressure (CPAP) Martina Mosing, Ulrike Auer, Paul MacFarlane, David Bardell, Johannes P. Schramel, Stephan H. Böhm, Regula Bettschart-Wolfensberger, Andreas D. Waldmann PII:
S1467-2987(17)30241-6
DOI:
10.1016/j.vaa.2017.06.004
Reference:
VAA 180
To appear in:
Veterinary Anaesthesia and Analgesia
Received Date: 22 March 2017 Revised Date:
11 May 2017
Accepted Date: 15 June 2017
Please cite this article as: Mosing M, Auer U, MacFarlane P, Bardell D, Schramel JP, Böhm SH, Bettschart-Wolfensberger R, Waldmann AD, Regional ventilation distribution and dead space in anaesthetised horses treated with and without continuous positive airway pressure (CPAP), Veterinary Anaesthesia and Analgesia (2017), doi: 10.1016/j.vaa.2017.06.004. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT RESEARCH PAPER Regional ventilation distribution and dead space in anaesthetised horses treated with and without continuous positive airway pressure (CPAP)
RI PT
Novel insights by electrical impedance tomography and volumetric capnography
Martina Mosinga,b, Ulrike Auerc, Paul MacFarlaned, David Bardelle, Johannes P
SC
Schramelc, Stephan H Böhmf, Regula Bettschart-Wolfensbergera & Andreas D
a
M AN U
Waldmannf
Section of Anaesthesiology, Vetsuisse Faculty, University of Zurich, Zurich,
Switzerland b
Veterinary University Vienna, Vienna, Austria
d
TE D
c
College of Veterinary Medicine, Murdoch University, Perth, Australia
Langford Veterinary Services, University of Bristol, Bristol, UK
e
School of Veterinary Clinical Science, University of Liverpool, Neston, UK
f
AC C
EP
Swisstom AG, Landquart, Switzerland
Correspondence
Martina Mosing, College of Veterinary Medicine, School of Veterinary and Life Sciences, Murdoch University, 90 South Street, 6150 Perth, Western Australia Tel: +61 8 9360 2641; E-mail:
[email protected]
Running head EIT and VCap during CPAP in horses
ACCEPTED MANUSCRIPT Acknowledgements and Conflict of interest statement We want to thank the “Stiftung Forschung für das Pferd” for financing this project. None of the authors of this paper has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of this
RI PT
paper.
Authors’ contributions
SC
All authors contributed to conception and design of research; MM, UA,
PM, RB, DB performed experiments; ADW, UA, SHB and JPS analysed data and did
M AN U
statistical analysis; all authors interpreted results; MM, UA, JPS drafted manuscript;
AC C
EP
TE D
all authors edited, revised and approved final version of the manuscript.
ACCEPTED MANUSCRIPT 1
Regional ventilation distribution and dead space in anaesthetized horses treated
2
with and without continuous positive airway pressure (CPAP).
3 Abstract
5
Objective The aim of this study was to evaluate the effect of continuous positive airway
6
pressure (CPAP) on regional distribution of ventilation and dead space in anaesthetised
7
horses.
8
Study design Randomized, experimental, crossover study.
9
Animals Eight healthy adult horses.
SC
RI PT
4
Methods Horses were anaesthetised twice with isoflurane in 50% oxygen and
11
medetomidine as continuous infusion in dorsal recumbency, and administered in
12
random order either CPAP (8 cmH2O) or NO CPAP for 3 hours. Electrical impedance
13
tomography (EIT) and volumetric capnography (VCap) measurements were performed
14
every 30 minutes. Lung regions with little ventilation [dependent (DSS) and non-
15
dependent silent spaces (NSS)], centre of ventilation (CoV) and dead space variables as
16
well as venous admixture were calculated. Statistical analysis was performed using
17
MANOVA and Pearson correlation.
18
Results Data from six horses were statistically analysed. In CPAP the CoV shifted to
19
dependent parts of the lungs (P <0.001) and DSS were significantly smaller (P<0.001),
20
while no difference was seen in NSS. Venous admixture was significantly correlated
21
with DSS with the treatment time taken as covariate (P <0.0001; r=0.65). No
22
differences were found for any VCap parameters.
23
Conclusions and clinical relevance In dorsally recumbent anaesthetised horses, CPAP
24
of 8 cmH2O results in redistribution of ventilation towards the dependent lung regions
25
thereby improving ventilation-perfusion matching. This improvement was not
AC C
EP
TE D
M AN U
10
ACCEPTED MANUSCRIPT 26
associated with an increase in dead space indicative for a lack in distension of the
27
airways or impairment of alveolar perfusion.
28 Keywords centre of ventilation, EIT, electrical impedance tomography, equine, silent
30
space, spontaneous ventilation
RI PT
29
AC C
EP
TE D
M AN U
SC
31
ACCEPTED MANUSCRIPT Introduction
33
General anaesthesia and change in body position results in a reduction of functional
34
residual capacity (FRC) of the lung to a level approximating that of residual volume
35
(Wahba 1991). This phenomenon promotes small airway closure, particularly in
36
dependent lung regions. Alveolar atelectasis develops distal to these closed airways as a
37
result of continued gas absorption, leading to lung regions with low ventilation /
38
perfusion ratios. In anaesthetised horses this is further compounded by compression of
39
lung tissue by the weight of the abdominal organs (Nyman et al. 1990).
SC
RI PT
32
40
Continuous positive airway pressure (CPAP) is a ventilation mode where airway
42
pressure is kept above ambient pressure throughout inspiration and expiration in
43
spontaneously breathing patients. The positive pressure applied to the lungs is expected
44
to keep small airways and alveoli open by maintaining FRC. Two recent studies have
45
demonstrated CPAP decreases venous admixture in anaesthetised horses and improves
46
gas exchange (Mosing et al. 2012b; Mosing et al. 2016a). Ventilation was not affected
47
by the airway pressure applied, indicated by unchanged end-tidal (PE´CO2) and arterial
48
partial pressure of carbon dioxide (PaCO2) (Mosing et al. 2016a). However, CPAP did
49
not prevent hypoventilation and PaCO2 levels were unacceptably high in both groups.
50
Half of the horses (in both groups) required controlled mechanical ventilation over the 6
51
hour study period.
TE D
EP
AC C
52
M AN U
41
53
Electrical impedance tomography (EIT) is a non-invasive monitoring modality that
54
provides imaging of regional lung function (Frerichs et al. 2016) using an array of
55
electrodes placed around the thorax. A small current is applied sequentially via two
56
electrodes while the voltage is measured by the remaining, allowing calculation of
ACCEPTED MANUSCRIPT impedance changes resulting from the varying gas and fluid content in the thorax. The
58
impedance map gained from these measurements can be analysed to elucidate regional
59
changes in ventilation distribution within the lung. Electrical impedance tomography
60
therefore offers new investigative possibilities to compare ventilation modes
61
particularly in horses where computed tomography is not possible (Mosing et al. 2016b;
62
Ambrisko et al. 2017). In humans it has been shown that CPAP causes a redistribution
63
of tidal ventilation towards dependent regions of the lungs in healthy subjects in supine
64
position (Andersson et al. 2011).
SC
RI PT
57
65
Volumetric capnography (VCap) enables the calculation of airway, alveolar and
67
physiologic dead space values by generating a plot of the expiratory partial pressure of
68
CO2 against expired tidal volume (Tusman et al. 2009). Using a custom-made flow-
69
partitioning device for large animals, positioned between the endotracheal tube and the
70
Y-piece of the anaesthetic breathing system, VCap and spirometry data can be measured
71
by standard human mainstream capnography sensors by recalculating the volume
72
measurements (Schramel et al. 2014; Ambrisko et al. 2017). Dead space measurements
73
have been suggested as a bedside method to determine the appropriate level of CPAP
74
and positive end-expiratory pressure (PEEP) consistent with keeping the small airways
75
open without over distending them (Spalding et al. 1999; Blankman et al. 2016).
TE D
EP
AC C
76
M AN U
66
77
The aim of this study was to evaluate regional ventilation distribution and dead space
78
parameters in anaesthetised horses treated with a CPAP level of 8 cmH2O and without
79
CPAP. We hypothesised that CPAP would redistribute ventilation towards the
80
dependent parts of the lungs without overdistending non-dependent lung regions.
81
ACCEPTED MANUSCRIPT Material and methods
83
This experimental prospective cross-over study was performed with the ethical approval
84
of the local committee for animal experimentation of the Swiss government
85
(TV-4985). This study was part of an original study performed to evaluate the effects of
86
CPAP on venous admixture, cardiac output and oxygen delivery in horses (Mosing et
87
al. 2016a) and the sample size was based on paramaters from the original study.
RI PT
82
88 Animals
90
Eight healthy adult horses from the original study (Mosing et al. 2016a) were included
91
in this study. Horses were considered healthy based on clinical examination, routine
92
haematology, biochemistry and arterial blood gas analysis before anaesthesia. They
93
were fasted for 12 hours, but had free access to water until 2 hours prior to induction.
94
Each horse was anaesthetised twice for 6 hours with one week rest and randomly (using
95
opaque envelopes) allocated to either administration of NO CPAP (anaesthetic machine
96
set to 0 cmH2O CPAP) or CPAP (continuous positive airway pressure of 8 cmH2O)
97
during the first anaesthesia. During the second anaesthesia the alternative protocol was
98
applied.
EP
AC C
99
TE D
M AN U
SC
89
100
Pre-experimental preparation and instrumentation
101
The day before anaesthesia a narrow 5 cm wide circumferential strip of hair was clipped
102
around the thorax at the level of the fifth to sixth intercostal space behind the elbow
103
joint to allow optimal electrical contact of the EIT electrodes with the skin. This
104
clipping also guaranteed the correct position of the belt during attachment in dorsal
105
recumbency during the first and second anaesthesia to guarantee comparable EIT
106
results.
ACCEPTED MANUSCRIPT 107
On the day of the experiment a jugular catheter (SecalonT; Ohmeda, Switzerland) and a
108
pulmonary artery catheter (8F 110 cm angiography balloon catheter; Arrow Swiss,
109
Switzerland and Intro-flex, Edwards Lifesciences, Switzerland) were placed using
110
pressure guidance to determine correct positioning.
RI PT
111 Anaesthesia
113
Horses were premedicated with medetomidine (0.007 mg kg-1) (Dorbene; Graeub AG,
114
Switzerland) IV and phenylbutazone (4 mg kg-1) IV (Butadion; Streuli Pharma AG,
115
Switzerland). Anaesthesia was induced with diazepam (0.02 mg kg-1) (Valium; Roche,
116
Switzerland) and ketamine (2 mg kg-1) (Ketanarkon 100; Streuli Pharma AG,
117
Switzerland) IV. The trachea of all horses were intubated using a 26 mm internal
118
diameter cuffed silicone tube. After placing the horse in dorsal recumbency, the
119
endotracheal tube was connected to a circle breathing system (Tafonius; Vetronic
120
Services LTD, UK). Isoflurane (Attane Isoflurane; Provet, Switzerland) vaporised in an
121
oxygen/air mix (inspiratory oxygen fraction: 0.5) in combination with a medetomidine
122
continuous rate intravenous infusion (CRI) (0.0035 mg kg-1 hour-1) was used to maintain
123
anaesthesia. A ketamine bolus (50-100 mg) was administered IV in case of spontaneous
124
movement. Anaesthesia was performed by the same experienced anaesthetist
125
throughout the experiments.
126
Ringer’s lactate (Ringer-Lactat Fresenius; Fresenius Kabi, Switzerland) was infused at a
127
rate of 10 mL kg-1 hour-1 and after two hours an infusion of hydroxyethyl starch (HAES-
128
steril 10% ad us. vet.; Fresenius Kabi, Switzerland) was started (1 mL kg-1 hour-1).
129
Hydroxyethyl starch was added to the standard cristalloid infusion to compensate for
130
the high urinary output expected with a medetomidine CRI (Mosing et al. 2016a). If
131
mean arterial pressure dropped below 75 mmHg dobutamine was infused intravenously
AC C
EP
TE D
M AN U
SC
112
ACCEPTED MANUSCRIPT initially at 0.03 mg kg-1 hour-1, increasing or decreasing by 0.006 mg kg-1 hour-1 steps
133
every 5 minutes to maintain mean arterial pressure between 75 and 85 mmHg.
134
If arterial partial pressure of CO2 (PaCO2) exceeded 100 mmHg (13.3 kPa) at any
135
measurement point, horses were excluded from further data acquisition and assisted
136
mechanical ventilation was initiated (Fig. 2).
137
Anaesthesia was maintained in all horses for six hours. However, for this study only the
138
first three hours of each anaesthesia were analysed as too many horses dropped out in
139
both treatment groups after that time point because of hypoventilation and the
140
requirement for mechanical ventilation. After the 6 hours of anaesthesia, horses
141
recovered unassisted in a padded recovery box.
142
M AN U
SC
RI PT
132
Monitoring and data collection
144
After positioning the horse in dorsal recumbency on an air mattress, electrically non-
145
conductive ultrasound gel was applied to the previously clipped region of the thorax.
146
Thereafter a custom-made EIT electrode belt was passed under and around the thorax.
147
The EIT Pioneer-Set (Swisstom AG, Switzerland) was used to acquire EIT data at a rate
148
of 46 frames per second. A modified Graz consensus reconstruction algorithm for EIT
149
(GREIT) (Adler et al. 2009) adapted to the horse’s anatomy was used to generate EIT
150
images for each horse representing breathing-related regional changes of impedance
151
(∆Z). Further details on EIT technology and image reconstruction can be found
152
elsewhere (Costa et al. 2008).
AC C
EP
TE D
143
153 154
The flow-partitioning device consisted of two identical flow splitter adapters – one
155
connected to the endotracheal tube and the other to the Y-piece of the anaesthetic
156
breathing system. Each adapter partitioned and merged the airflow to four equal parts
ACCEPTED MANUSCRIPT and directed the flow through four identical human adult connectors of a combined
158
mainstream CO2 infrared and differential pressure sensor (NICO, Respironics,
159
Wallingford, CT, USA) via transparent silicone tubes. Further details on the flow-
160
partitioning device can be found elsewhere (Schramel et al. 2014). The CO2 mainstream
161
sensor and the spirometer interface were connected to one of the four NICO sensors.
162
Volumetric capnography and spirometry data were recorded using dedicated software
163
Datacoll (Respironics; Wallingford, CT, USA).
165
Electrical impedance tomography, VCap and spirometry data were recorded over three
166
minutes every 30 minutes during the first three hours of anaesthesia (M30, M60, M90,
167
M120, M150, M180).
M AN U
SC
164
RI PT
157
168
At each measurment point arterial and mixed venous blood samples were taken
170
simultaneously from the facial and pulmonary artery catheters under anaerobic
171
conditions and PO2, PCO2 and haemoglobin concentration (Hb) were analysed
172
immediately (Rapidpoint; Siemens, Switzerland).
TE D
169
EP
173
The blood gas analyser and NICO were calibrated following manufacturers’ guidelines
175
before each experiment.
176
AC C
174
177
Standard cardiopulmonary monitoring was employed using multi-parameter monitors
178
(Datex-Ohmeda S/3 Anaesthesia Monitor; Datex-Ohmeda and Tafonius). Values were
179
recorded manually every five minutes and are reported elsewhere (Mosing et al.,
180
2016a).
181
ACCEPTED MANUSCRIPT Data Analysis
183
The EIT regions of interest (ROI) for the right and left lung were determined by
184
creating individualized lung contours for each horse based on the assumption that the
185
large ventilation-induced impedance changes (∆Z) within the lungs will exceed by far
186
those of other thoracic structures. Therefore, ∆Z values of more than 10% of the
187
maximal impedance amplitude in any of the recorded breaths were used to define the
188
outer boundaries of the ‘EIT-based lung region’ for each individual. This means that
189
every pixel showing an amplitude of more than 10% during a breath was included in the
190
individual ROI. Only pixels within these ROIs were analysed. At each time point, tidal
191
EIT images were created from 10 consecutive stable breaths by calculating the
192
impedance difference between inspiration and the preceding end-expiration for each
193
pixel.
194
The Centre of ventilation (CoV; Fig. 1) was determined as a percentage of ventro-dorsal
195
extension of the lung region, where 0% referred to ventilation occurring in the most
196
ventral part of the lung and 100% to that in the most dorsal part (Frerichs et al. 1998;
197
Mosing et al. 2016b) (Fig. 1). For dorsal recumbency this means that values < 50%
198
represented predominant ventilation of the non-dependent parts, while values > 50 %
199
indicated a CoV in dependent parts of the lungs. For each breath a virtual line was
200
drawn horizontally through the CoV defined as the ventilation horizon.
201
Silent spaces indicate lung regions with < 10% impedance change of the maximum
202
within the tidal image and are expressed as percentages of the total lung area (Ukere et
203
al 2016). Dependent silent spaces (DSS) lie below the ventilation horizon and indicate
204
collapsed or poorly ventilated lung regions, while non-dependent silent spaces (NSS)
205
are located above the ventilation horizon and indicate over-distended lung regions
206
(Ukere et al. 2016) (Fig. 1).
AC C
EP
TE D
M AN U
SC
RI PT
182
ACCEPTED MANUSCRIPT 207 208
Figure 1 near here
209 Raw CO2 and flow data recorded by the dedicated NICO software were used to evaluate
211
spirometry data and construct VCap curves for all or a maximum of 10 breaths recorded
212
over the 3 minutes. The latter was computed with a custom-made macro routine in
213
Excel (Excel; Microsoft Corporation, WA, USA). The curve-fitting algorithm was
214
realized by the solver function according to the formula developed by Tusman and
215
colleagues (2009).
SC
RI PT
210
M AN U
216
The following spirometry, capnography and VCap parameters were analysed:
218
Tidal volume (VT), respiratory rate (fR), end-tidal PCO2 (PE´CO2), Bohr´s dead space
219
ratio (VDBohr), Bohr-Enghoff’s dead space ratio (VDB-Eng), airway dead space to tidal
220
volume ratio (VDaw/VT), alveolar dead space (calculated from VDB-Eng) to alveolar tidal
221
volume ratio (VDalv/VTalv) and volume of expired CO2 per breath (VCO2,br).
TE D
217
222
Venous admixture (Qs/Qt) was calculated retrospectively using the standard equation
224
(Lumb 2010): Qs/Qt = (Cc'O2-CaO2)/(Cc’O2- CvO2). Details on the equation and
225
absolute values are reported elsewhere (Mosing et al., 2016a).
AC C
226
EP
223
227
Statistics
228
Analysis was performed using NCSS statistical software, (NCSS V 11.0.3, Statistical
229
Software (2016). NCSS, LLC. Kaysville, Utah, USA, ncss.com/software/ncss.) Data
230
were assessed for normality by Kolmogorov-Smirnov-test. Data of CPAP and NO
231
CPAP were analyzed with MANOVA in regards to influence of treatment and time on
ACCEPTED MANUSCRIPT 232
results, followed by post-hoc Bonferroni test. Pearson correlation was calculated for
233
venous admixture and DSS. Statistical significance was set at P< 0.05. Results are given
234
as mean values ± standard deviation and range.
AC C
EP
TE D
M AN U
SC
RI PT
235
ACCEPTED MANUSCRIPT Results
237
A total of eight horses were enrolled in this study. Data from two horses was excluded
238
because of technical difficulties with the EIT device in one horse and with volumetric
239
capnography data collection in the other (Fig. 2). Therefore, data of six horses were
240
included in the statistical analysis. Only the first three hours of anaesthesia were
241
analysed because of the small number of horses remaining in both groups after this time
242
point (Fig. 2). The six horses had an age of 10 ± 6 years and body weight 539 ± 35 kg (1
243
Thoroughbred, 5 Freiberger).
244
One and two horses required mechanical ventilation because of high PaCO2 levels after
245
M150 in group CPAP and NO CPAP, respectively. One horse in group CPAP moved
246
after M60 and did not regain spontaneous ventilation for more than 10 minutes after
247
administration of a bolus of ketamine. Data from this horse was excluded after M60
248
(Fig. 2).
249
The airway pressure swings during inspiration and expiration for these spontaneously
250
breathing horses were 8 ± 8 cmH2O and 0 ± 10 cmH2O for CPAP and NO CPAP,
251
respectively.
252
Figure 2 near here
SC
M AN U
TE D
EP
AC C
253
RI PT
236
254
Respiratory variables, VCap and EIT data of the six horses are given in Table 1.
255
Dependent silent spaces were significantly lower in CPAP (P<0.001), while no
256
difference was seen in NSS. The CoV was higher in CPAP, located more dorsally, in
257
the dependent parts of the lungs (P<0.001). Venous admixture was significantly
258
correlated with DSS when time was taken as covariate (P <0.0001; r=0.65).
ACCEPTED MANUSCRIPT 259
No difference was found for VT (P=0.932), fR (P=0.986), PE´CO2 (P=0.981), VDaw/VT
260
(P=0.456), VDalv/VTalv (P=0.955), VDBohr (P=0.835), VDB-Eng (P=0.456) or VCO2,br
261
(P=0.971) between the two groups.
262 Table 1 near here
RI PT
263
AC C
EP
TE D
M AN U
SC
264
ACCEPTED MANUSCRIPT Discussion
266
This study demonstrates that CPAP of 8 cmH2O in dorsally recumbent anaesthetised
267
horses decreases the amount of silent spaces in the dependent parts of the lungs. This
268
results in an improved matching of ventilation and perfusion, seen as a decrease in
269
venous admixture. At the given airway pressure CPAP did not cause an increase in any
270
dead space parameters neither overdistending the airways (no change in VDaw/VT) nor
271
the alveoli (no change in VDalv/VTalv).
SC
272
RI PT
265
Electrical impedance tomography is a radiation free, breath-by-breath functional
274
imaging modality, which can be used to monitor regional lung ventilation in different
275
species (Frerichs et al. 2016; Ambrisko et al. 2017; Mosing et al. 2017) and has been
276
used in humans to verify the efficacy of CPAP (Andersson et al. 2011). In awake supine
277
humans as well as in our anaesthetised horses in dorsal recumbency a redistribution of
278
tidal ventilation towards dorsal (dependent) parts of the lungs was found with
279
administration of CPAP (Andersson et al. 2011). In the horse - as it is in all quadrupeds,
280
lung perfusion is considered higher in dorsal parts of the lungs unrelated to posture
281
(Hlastala et al. 1996). This is even more marked in anaesthetised horses (Dobson et al.
282
1985). The shift of the distribution of ventilation into the well perfused dorsal parts of
283
the lungs caused by CPAP in our horses lead to an improvement in ventilation /
284
perfusion matching and gas exchange indicated by lower venous admixture in CPAP.
285
Both, induction of general anaesthesia and dorsal recumbency reduce functional
286
residual capacity (FRC) and cause collapse of the small airways (Hedenstierna &
287
Edmark 2010; Spaeth et al. 2016). This decrease in FRC is because of the loss of
288
respiratory muscle tone and compression by viscera, causing a marked decrease in lung
289
volume. Applying CPAP increases FRC and counteracts airway collapse and therefore
AC C
EP
TE D
M AN U
273
ACCEPTED MANUSCRIPT atelectasis formation (Edmark et al. 2014). Dependent silent spaces of the EIT represent
291
poorly ventilated areas in the dependent parts of the lungs and correspond to collapsed
292
or poorly ventilated lung areas (Ukere et al. 2016). We found a significantly lower
293
percentage of DSS when CPAP was applied to the lungs of the horses. This confirms
294
that CPAP reduces lung collapse, probably by keeping small airways patent in the
295
dependent lung areas. We furthermore found a direct correlation between the decrease
296
in DSS and venous admixture over time, which agrees with previous findings in humans
297
that CPAP not only improves V/Q matching but also reduces venous admixture by
298
counteracting airway collapse and formation of atelectasis (Edmark et al. 2014).
SC
RI PT
290
M AN U
299
It has been previously reported (Mosing et al. 2016b; Ambrisko et al. 2017) that the
301
effects of controlled mechanical ventilation (CMV) are different from those of CPAP in
302
dorsal recumbent anaesthetised horses. CPAP and lung recruitment manoeuvres shift
303
the CoV towards the better perfused dependent parts of the lungs, while just switching
304
from spontaneous ventilation to CMV moved the CoV towards the ventral non-
305
dependent parts (Mosing et al. 2016b; Ambrisko et al. 2017). This ventral shift was
306
explained by the loss of dorsal movement of the diaphragm when switching from
307
spontaneous ventilation to CMV (Froese & Bryan 1974; Mosing et al. 2016b). During
308
spontaneous ventilation, the dorsal dependent parts of the diaphragm show greater
309
movement than the ventral non-dependent parts thereby drawing the inflowing gas
310
towards the dependent parts. When switching to CMV the diaphragm is pushed
311
caudally to the same extent in the ventral and dorsal parts, which causes the ventilation
312
to shift ventrally (Radke et al. 2012; Mosing et al. 2016b). Performing a recruitment
313
manoeuvre initially shifted the CoV towards the dorsal dependent parts of the lungs
314
during CMV possibly as a result of opening atelectasis at those sites (Ambrisko et al.
AC C
EP
TE D
300
ACCEPTED MANUSCRIPT 2017). However, because of the insufficient amounts of PEEP applied after the
316
manoeuvre, the CoV returned to baseline and therefore towards the ventral non-
317
dependent less perfused lung parts. Despite the relatively low CPAP level of 8 cmH2O
318
in our study, the CoV was located more dorsally throughout the study period with
319
CPAP.
RI PT
315
320
The volumetric capnogram is the dynamic representation of the CO2 elimination
322
plotting the partial pressure of CO2 over the exhaled volume. The best fit approximation
323
of the measurements by a mathematical model curve proposed by Tusman et al. (2009)
324
enables the calculation of different pulmonary dead space parameters.
325
This method can be applied to all mammals as long as the trachea and lung follow
326
similar architectures. A precondition for the measurement is the use of a mainstream
327
capnograph (i.e. a fast infrared) CO2 sensor and flow meter to avoid a time delay
328
between theses signals. The integration of a human mainstream airway adapter in a
329
special flow-partitioning device for large animals and mathematical correction for the
330
splitting of the tidal volume enables the calculation of VCap parameters also in large
331
animals such as horses (Moens et al. 2014; Sacks & Mosing 2016; Ambrisko et al.
332
2017). In recent papers, VCap variables have been shown to detect distension of the
333
lung tissue during a recruitment manoeuvre in horses (Ambrisko et al. 2017) and can be
334
used to distinguish between distension of the airways and overdistension of the alveolar
335
compartment during PEEP titration in humans (Blankman et al. 2016). Furthermore, it
336
has been used to detect optimal CPAP levels by observing changes in VDB-Eng (Spalding
337
et al. 1999). In our study, no differences in physiologic, airway or alveolar dead space
338
were found between groups. Overdistension of alveoli would be rather unphysiological
339
during spontaneous breathing and is unlikely with administration of CPAP in contrast to
AC C
EP
TE D
M AN U
SC
321
ACCEPTED MANUSCRIPT use of CMV or recruitment manoeuvres where high peak inspiratory pressures are used.
341
However, the continuous positive pressure in the alveoli can impair capillary blood flow
342
around the alveoli causing an increase in lung units with high V/Q mismatch or even an
343
increase in alveolar dead space. The small difference we found between CPAP and NO
344
CPAP in all dead space parameters suggests that a positive airway pressure of 8 cmH2O
345
neither causes airway distension nor interruption of alveolar perfusion in healthy
346
anaesthetised horses.
347
However, one would expect a higher airway dead space in CPAP because positive
348
pressures keep small airways open, especially the terminal and respiratory bronchioles,
349
the typical site of airway collapse. Airway dead space by definition is the volume of the
350
airways in which movement of gases occurs by convection (Fletcher et al. 1981). As gas
351
movement in terminal bronchioles takes place by diffusion and not convectional flow,
352
the volume of gas within this ‘space’ does not contribute to the volume of airway dead
353
space. As long as the alveoli corresponding to these terminal bronchioles are still
354
perfused, alveolar dead space will remain constant as well (Fletcher et al. 1981). This
355
again suggests that the CPAP level used in our study had no negative impact on gas
356
flow in the airways or blood flow in the capillaries.
SC
M AN U
TE D
EP
AC C
357
RI PT
340
358
Another evidence for similar pulmonary perfusion conditions in both groups is the small
359
difference in VCO2,br. The volume of expired CO2 per breath is a key VCap variable
360
and a very sensitive indicator of changes in pulmonary blood flow (Suarez-Sipman et al.
361
2016). It can also be used to identify the efficiency of ventilation (Tusman et al. 2012;
362
Suarez-Sipman et al. 2016). The small difference we found between the two groups for
363
VCO2,br confirms the finding of our previous study where CPAP had no influence on
364
the efficiency of ventilation per breath when compared to NO CPAP (Mosing et al.
ACCEPTED MANUSCRIPT 365
2016a). However, minute ventilation was insufficient in both groups leading to an
366
increase in PE´CO2 over time. For this reason, CMV might be necessary for the longer
367
clinical procedures even when applying CPAP.
368 One limitation of our study is that all findings are only related to a CPAP of 8 cmH2O.
370
Hence we cannot extrapolate them to lower or higher CPAP levels. This pressure was
371
chosen based on findings of a previous study showing that a CPAP of 8 cmH2O was the
372
minimal pressure necessary to maintain airway pressure above 0 cmH2O during the
373
entire breathing circle in anaesthetised horses in dorsal recumbency (Mosing et al.
374
2012b). However, based on the shifts in the CoV over time it is likely that a higher
375
airway pressures may have been needed to prevent collapse of the dependent lungs.
M AN U
SC
RI PT
369
376
Despite the limitation of only a small number of animals included in this study a
378
significant difference was found for EIT variables, while a very high agreement
379
between the two groups was seen for the dead space variables. However, too many
380
horses dropped out of the study after the first three hours of anaesthesia to allow for
381
statistical analysis after this time point. The benefits of CPAP on cardiorespiratory
382
variables were shown over 6 hours of anaesthesia in the same horses reported in an
383
earlier paper (Mosing et al., 2016a). It remains unknown if those benefits are reflected
384
by EIT and VCap variables after the first three hours of anaesthesia.
EP
AC C
385
TE D
377
386
Conclusion
387
In anaesthetised horses in dorsal recumbency a CPAP level of 8 cmH2O reduces
388
collapsed lung areas in the dependent dorsal part of the lungs and distributes ventilation
389
towards these regions thereby improving the matching of ventilation and perfusion. This
ACCEPTED MANUSCRIPT 390
is likely to be a result of maintained small airway patency without causing an increase
391
in airway or alveolar dead space.
AC C
EP
TE D
M AN U
SC
RI PT
392
ACCEPTED MANUSCRIPT References
394
Ambrisko TD, Schramel JP, Hopster K, et al. (2017) Assessment of distribution of
395
ventilation and regional lung compliance by electrical impedance
396
tomography in anaesthetized horses undergoing alveolar recruitment
397
manoeuvres. Vet Anaesth Analg, doi.org/10.1016/j.vaa.2016.03.001.
399 400
Adler A, Arnold JH, Bayford R et al. (2009) GREIT: a unified approach to 2D linear EIT reconstruction of lung images. Physiol Meas, 30, S35-55.
Andersson B, Lundin S, Lindgren S et al. (2011) End-expiratory lung volume and
SC
398
RI PT
393
ventilation distribution with different continuous positive airway pressure
402
systems in volunteers. Acta Anaesthesiol Scand, 55, 157-164.
M AN U
401
Blankman P, Shono A, Hermans BJ et al. (2016) Detection of optimal PEEP for equal
404
distribution of tidal volume by volumetric capnography and electrical
405
impedance tomography during decreasing levels of PEEP in post cardiac-
406
surgery patients. Br J Anaesth, 116, 862-869.
408
Costa EL, Chaves CN, Gomes S et al. (2008) Real-time detection of pneumothorax using electrical impedance tomography. Crit Care Med, 36, 1230-1238.
EP
407
TE D
403
Dobson A, Gleed RD, Meyer RE et al. (1985) Changes in blood flow distribution in
410
equine lungs induced by anaesthesia. Q J Exp Physiol, 70, 283-297.
411 412 413 414
AC C
409
Edmark L, Auner U, Hallen J et al. (2014) A ventilation strategy during general anaesthesia to reduce postoperative atelectasis. Ups J Med Sci, 119, 242-
250.
Fletcher R, Jonson B, Cumming G et al. (1981) The concept of deadspace with
415
special reference to the single breath test for carbon dioxide. Br J Anaesth,
416
53, 77-88.
ACCEPTED MANUSCRIPT 417
Frerichs I, Amato MB, van Kaam AH et al. (2016) Chest electrical impedance
418
tomography examination, data analysis, terminology, clinical use and
419
recommendations: consensus statement of the TRanslational EIT
420
developmeNt stuDy group. Thorax, doi:10.1136/thoraxjnl-2016- 208357 Frerichs I, Hahn G, Golisch W et al. (1998) Monitoring perioperative changes in
422
distribution of pulmonary ventilation by functional electrical impedance
423
tomography. Acta Anaesthesiol Scand, 42, 721-726.
426 427 428
SC
425
Froese AB, Bryan AC (1974) Effects of anesthesia and paralysis on diaphragmatic mechanics in man. Anesthesiology, 41, 242-255.
Hedenstierna G, Edmark L (2010) Mechanisms of atelectasis in the perioperative
M AN U
424
RI PT
421
period. Best Pract Res Clin Anaesthesiol, 24, 157-169. Hlastala MP, Bernard SL, Erickson HH et al. (1996) Pulmonary blood flow distribution in standing horses is not dominated by gravity. J Appl Physiol
430
(1985), 81, 1051-1061.
433 434 435 436
Oxford, UK, pp. 119-144
EP
432
Lumb AB (2010) Nunn's applied respiratory physiology (7th edn). Elsevier,
Moens Y, Schramel JP, Tusman G et al. (2014) Variety of non-invasive continuous monitoring methodologies including electrical impedance tomography
AC C
431
TE D
429
provides novel insights into the physiology of lung collapse and recruitment
- case report of an anaesthetized horse. Vet Anaesth Analg, 41, 196-205.
437
Mosing M, Iff I, Hirt R et al. (2012a) Evaluation of variables to describe the shape of
438
volumetric capnography curves during bronchoconstriction in dogs. Res Vet
439
Sci, 93, 386-392.
ACCEPTED MANUSCRIPT 440
Mosing M, MacFarlane P, Bardell D et al. (2016a) Continuous positive airway
441
pressure (CPAP) decreases pulmonary shunt in anaesthetized horses. Vet
442
Anaesth Analg. 43, 611-622
443
Mosing M, Marly-Voquer C, MacFarlane P et al. (2016b) Regional distribution of ventilation in horses in dorsal recumbency during spontaneous and
445
mechanical ventilation assessed by electrical impedance tomography: a
446
case series. Vet Anaesth Analg. doi: 10.1111/vaa.12405
Mosing M, Rysnik M, Bardell D et al. (2012b) Use of continuous positive airway
SC
447
RI PT
444
pressure (CPAP) to optimise oxygenation in anaesthetised horses--a clinical
449
study. Equine Vet J, 45, 414-418.
450
M AN U
448
Mosing M, Sacks M, Tahas SA et al. (2017) Ventilatory incidents monitored by
451
electrical impedance tomography in an anaesthetized orangutan (Pongo
452
abelii). Vet Anaesth Analg, in press.
455
TE D
454
Nyman G, Funkquist B, Kvart C et al. (1990) Atelectasis causes gas exchange impairment in the anaesthetised horse. Equine Vet J, 22, 317-324. Radke OC, Schneider T, Heller AR et al. (2012) Spontaneous breathing during
EP
453
general anesthesia prevents the ventral redistribution of ventilation as
457
detected by electrical impedance tomography: a randomized trial.
458 459 460 461 462
AC C
456
Anesthesiology, 116, 1227-1234.
Sacks M, Mosing M (2016) Volumetric capnography to diagnose venous air embolism in an anaesthetised horse. Vet Anaesth Analg.
doi: 10.1111/vaa.12383 Schramel JP, Wimmer K, Ambrisko TD et al. (2014) A novel flow partition device
463
for spirometry during large animal anaesthesia. Vet Anaesth Analg, 41, 191-
464
195.
ACCEPTED MANUSCRIPT 465
Spaeth J, Daume K, Goebel U et al. (2016) Increasing positive end-expiratory
466
pressure (re-)improves intraoperative respiratory mechanics and lung
467
ventilation after prone positioning. Br J Anaesth, 116, 838-846. Spalding HK, Banner MJ, Skimming JW (1999) Selection of an appropriate level of
469
continuous positive airway pressure (CPAP) using real time measurement
470
of physiologic dead space to tidal volume ratio (VD/VT). Critical Care
471
Medicine, 27, A106.
474
SC
473
Suarez-Sipmann F, Bohm SH, Tusman G (2014) Volumetric capnography. Cur Opin Crit Care, 20, 333–339.
Tusman G, Scandurra A, Bohm SH et al. (2009) Model fitting of volumetric
M AN U
472
RI PT
468
475
capnograms improves calculations of airway dead space and slope of phase
476
III. J Clin Monit Comput, 23, 197-206.
478
Tusman G, Sipmann FS, Bohm SH (2012) Rationale of dead space measurement by volumetric capnography. Anesth Analg, 114, 866-874.
TE D
477
Ukere A, Marz A, Wodack KH et al. (2016) Perioperative assessment of regional
480
ventilation during changing body positions and ventilation conditions by
481
electrical impedance tomography. Br J Anaesth, 117, 228-235.
483 484
Wahba RW (1991) Perioperative functional residual capacity. Can J Anaesth, 38,
AC C
482
EP
479
384-400.
ACCEPTED MANUSCRIPT 485 486
Figure legend
487 Figure 1 Left: Graph of a horse in dorsal recumbency with a projected electrical
489
impedance tomography (EIT) image. Right: Schematic illustration of the measured
490
variables within the EIT image and the two regions of interest. Centre of ventilation
491
(CoV), non-dependent (NSS) and dependent (DSS) silent space and ventilation horizon.
RI PT
488
SC
492
Figure 2 Flow diagram showing the number of horses included in this study and drop
494
out at each measurement point (M).
495
CPAP; continuous positive airway pressure.
M AN U
493
AC C
EP
TE D
496
ACCEPTED MANUSCRIPT
M60
M90
M120
M150
M180
CPAP
NO CPAP
CPAP
NO CPAP
CPAP
NO CPAP
CPAP
NO CPAP
CPAP
NO CPAP
CPAP
VT (L)
6.1±2.10
6.0±0.9
6.6±1.5
7.2±2.3
6.3±1.5
6.6±1.4
6.9±1.8
8.2±2.2
6.1±1.4
7.5±3.0
6.2±0.9
6.4±1.3
fR (breaths minute-1)
7±6
7±5
6±3
6±4
6±3
6±4
6±3
6±4
6±3
6±4
7±2
6±3
PE’CO2 (kPa)
7.0±0.5
6.8±0.7
7.6±0.9
7.3±0.7
8.0±1.2
7.7±0.9
8.2±1.3
8.0±1.2
8.2±0.9
8.1±1.2
8.4±1.1
8.0±0.9
VDBohr
0.37±0.1
0.41±0.06
0.36±0.1
0.36±0.07
0.37±0.1
0.34±0.09
0.36±0.1
0.32±0.06
0.38±0.05
0.35±0.07
0.38±0.04
0.37±0.06
VDB-Eng
0.54±0.05
0.55±0.05
0.51±0.03
0.53±0.04
0.51±0.05
0.54±0.05
0.56±0.05
0.50±0.05
0.59±0.06
0.54±0.06
0.53±0.06
0.57±0.05
VDaw/VT
0.36±0.08
0.39±0.08
0.34±0.06
0.33±0.08
0.35±0.06
0.32±0.09
0.32±0.07
0.29±0.06
0.36±0.08
0.33±0.08
0.33±0.05
0.35±0.08
VDalv/VTalv
0.26±0.07
0.27±0.07
0.29±0.06
0.26±0.06
0.30±0.11
0.25±0.06
0.30±0.05
0.35±0.06
0.31±0.05
0.35±0.06
0.33±0.02
0.29±0.15
VCO2,br (mL)
279±131
264±66
326±118
356±197
303±99
350±157
356±99
461±199
331±86
431±309
283±56
324±116
CoV (%)*
51.6±2.01
55.2±4.1
51.3±3.1
54.8±2.5
52.5±1.5
55.0±1.6
52.3±1.6
55.3±0.7
54.6±1.8
56.4±1.2
55.2±0.9
56.5±0.9
NSS (%)
12.8±2.5
12.3±7.9
10.2±5.3
12.1±5.2
12.2±1.7
11.5±4.5
11.7±3.6
13.3±3.6
14.8±3.6
14.3±5.0
17.4±2.4
14.8±2.7
DSS (%)*
16.6±6.4
6.8±5.5
12.8±4.8
8.3±2.8
13.7±5.3
8.2±2.8
13.01±4.5
8.3±3.7
11.8±3.6
6.7±4.5
11.5±4.6
7.0±2.6
Venous admixture (%)*
23.1±12.6
11.0±4.6
25.4±12.6
15.4±8.3
35.7±11.6
28.9±14.3
30±11.4
23.3±11.3
33.5±13.6
24.1±13.0
34.4±12.6
27.5±7.2
AC C
EP
TE D
M AN U
NO CPAP
SC
M30
RI PT
Table 1: Mean ± standard deviation of respiratory, volumetric capnography and electrical impedance tomography EIT parameters in horses without positive airway pressure (NO CPAP) or administered continuous positive airway pressure of 8 cmH2O (CPAP) at different time points (30 – 180 min after induction = M30 – 180).
VT, tidal volume; fR, respiratory rate; PE’CO2, end-tidal carbon dioxide; VDBohr, Bohr’s dead space ratio; VDB-Eng, Bohr-Enghoff’s dead space ratio; VDaw/VT, airway dead space per tidal volume; VDalv/VTalv, alveolar dead space as a ratio to alveolar tidal volume; VCO2,br, volume of expired CO2 per
ACCEPTED MANUSCRIPT
breath; CoV, centre of ventilation; NSS, non-dependent silent space; DSS, dependent silent space and venous admixture (%). *Significant difference
AC C
EP
TE D
M AN U
SC
RI PT
between groups (p < 0.05)
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT